Storytelling in visualisation

There’s that old cliche adage of “telling a story with data”.

You intro your story with a description of the data-related problem. Usually top-level and in a way that the reader is clearly bought into the premise.

Then you bulk out the middle of your story with explanations for how you began to tackle the question - you looked at these metrics, extrapolated them over time, saw some patterns and dove deeper until finally you announce the exciting conclusion of

your findings. Maybe you bold this, or pause for dramatic effect.

Then finally you finish off your story with a recommendation or solution, perhaps something you’ve already addressed, or something your stakeholder will need to raise with another team to resolve.

This clearly works in a written or verbal setting. I’ve done it. We’ve all done it. Of course accompanied with charts or data exports.

Looker Studio dashboard.

Data visualisation can do the same thing. Ensuring a data dashboard adheres to a storytelling-archetype, can help to keep stakeholder eyeballs on your dashboard for longer.

Firstly I want to mention language. Keep things simple. Really simple. It’s no coincidence that some of the most successful stories ever written have been primarily for children. Adults yearn for simplicity in a complicated world, and while you want your viewer to be engaged, you’ll also want them to understand at first glance everything your dashboard has to offer.

Now, no two dashboards are ever the same. However, all data needs to be presented with some top level context. Typically a viewer will read a dashboard from top to bottom, usually from left to right. Therefore the first context-building metric you’ll want your stakeholder to see should be in the top-left of the dashboard. This can be accompanied with other top-level metrics (or key KPIs), which will set the scene nicely for what’s to follow.

Scorecard widgets from a Looker Studio dashboard.

Next you’ll want to introduce some filters or breakdowns to these KPIs. How are these broken down over time, or how do these look when applying specific dimension filters to them? You’re now starting to begin leading the stakeholder down your own train of thought as an analyst - as a storyteller - as you’re setting the scene and invoking their intrigue.

Line charts and tables from a Looker Studio dashboard.

Now you’ll want to hit them with your main character. The reason the stakeholder has clicked to open your dashboard. The KPI that they “really” care about. Except that they haven’t just been “told a number”, they see that number, or value, in the context of overall performance. It immediately has meaning, and any immediate questions will have already been answered.

Your main character KPI will need to be shown off with all manner of tables/graphs/charts that explain why that KPI value is that value. Is it because of X sales, Y forms filled, or Z bookings? Be succinct, but also generous. This is where company-wide insight generation can be formed, and where further questions can be raised throughout the business.

Table, column and bar chart from a Looker Studio dashboard.

Beyond this, as an analyst, is where your own experience and imagination can take flight. What else do customers do on your product or website? Identified as secondary KPIs, as with many stories, the heroism of your main character can only be weighed against the performance of their friends. A submitted KMI can lead to a future sale, or watched video can be the hook a customer needs to feel commitment.

To caveat, this concept won’t always make sense with every data viz project. However, in many cases, applying a storytelling approach to visualising data can help make it more interesting, more compelling, and will keep the reader coming back for more.